src/streamlit_app.py

import streamlit as st
import requests
import base64
import os
from typing import List, Dict, Tuple, Optional
import json

# Must be the first Streamlit command
st.set_page_config(
    page_title="GitHub Repository Tree Generator",
    page_icon="🌳",
    layout="wide",
    initial_sidebar_state="expanded"
)

class GitHubRepoAnalyzer:
    def __init__(self):
        self.github_token = os.environ.get('GITHUB_TOKEN')
        self.headers = {
            'Accept': 'application/vnd.github.v3+json',
            'User-Agent': 'GitHubRepoAnalyzer'
        }
        if self.github_token:
            self.headers['Authorization'] = f'token {self.github_token}'
    
    def parse_github_url(self, url: str) -> Tuple[str, str]:
        """Parse GitHub URL to extract owner and repo name."""
        url = url.strip()
        if url.endswith('/'):
            url = url[:-1]
        
        if 'github.com/' in url:
            parts = url.split('github.com/')[-1].split('/')
            if len(parts) >= 2:
                owner = parts[0]
                repo = parts[1]
                if repo.endswith('.git'):
                    repo = repo[:-4]
                return owner, repo
        
        raise ValueError("Invalid GitHub URL format")
    
    def get_branches(self, owner: str, repo: str) -> List[str]:
        """Fetch available branches for the repository."""
        try:
            url = f"https://api.github.com/repos/{owner}/{repo}/branches"
            response = requests.get(url, headers=self.headers, timeout=10)
            
            if response.status_code == 200:
                branches = response.json()
                return [branch['name'] for branch in branches]
            else:
                return ['main', 'master']
        except Exception:
            return ['main', 'master']
    
    def get_repo_tree(self, owner: str, repo: str, branch: str = 'main', subfolder: str = '') -> Dict:
        """Get the complete tree structure of a repository."""
        try:
            url = f"https://api.github.com/repos/{owner}/{repo}/git/trees/{branch}?recursive=1"
            response = requests.get(url, headers=self.headers, timeout=30)
            
            if response.status_code == 404:
                url = f"https://api.github.com/repos/{owner}/{repo}/git/trees/master?recursive=1"
                response = requests.get(url, headers=self.headers, timeout=30)
            
            if response.status_code != 200:
                return {"error": f"Failed to fetch repository tree. Status: {response.status_code}"}
            
            tree_data = response.json()
            
            filtered_tree = []
            for item in tree_data.get('tree', []):
                path = item['path']
                if subfolder:
                    if not path.startswith(subfolder):
                        continue
                
                filtered_tree.append({
                    'path': path,
                    'type': item['type'],
                    'size': item.get('size', 0),
                    'url': item.get('url', '')
                })
            
            return {
                "owner": owner,
                "repo": repo,
                "branch": branch,
                "subfolder": subfolder,
                "tree": filtered_tree
            }
        
        except Exception as e:
            return {"error": f"Error fetching repository tree: {str(e)}"}
    
    def format_tree_display(self, tree_data: Dict) -> str:
        """Format tree data for display."""
        if "error" in tree_data:
            return f"❌ {tree_data['error']}"
        
        tree = tree_data.get('tree', [])
        if not tree:
            return "📁 No files found in the specified path."
        
        # Header information
        output = []
        output.append(f"📦 Repository: {tree_data['owner']}/{tree_data['repo']}")
        output.append(f"🌿 Branch: {tree_data['branch']}")
        if tree_data['subfolder']:
            output.append(f"📁 Subfolder: {tree_data['subfolder']}")
        output.append(f"📊 Total items: {len(tree)}")
        output.append("\n" + "="*60 + "\n")
        
        # Build tree structure
        tree_structure = self._build_tree_structure(tree, tree_data.get('subfolder', ''))
        output.append(tree_structure)
        
        return "\n".join(output)
    
    def _build_tree_structure(self, tree_items: List[Dict], subfolder: str = '') -> str:
        """Build a hierarchical tree structure with proper indentation and lines."""
        if not tree_items:
            return ""
        
        # Create a directory structure
        dir_structure = {}
        
        # Process all items and build directory structure
        for item in tree_items:
            path = item['path']
            
            # Remove subfolder prefix if present
            if subfolder and path.startswith(subfolder):
                path = path[len(subfolder):].lstrip('/')
            
            if not path:  # Skip empty paths
                continue
                
            parts = path.split('/')
            current = dir_structure
            
            # Build nested structure
            for i, part in enumerate(parts):
                if part not in current:
                    current[part] = {
                        '_type': 'tree' if i < len(parts) - 1 or item['type'] == 'tree' else 'blob',
                        '_size': item.get('size', 0) if i == len(parts) - 1 else 0,
                        '_children': {}
                    }
                current = current[part]['_children']
        
        # Generate tree display
        return self._format_tree_recursive(dir_structure, "", True)
    
    def _format_tree_recursive(self, structure: Dict, prefix: str = "", is_root: bool = False) -> str:
        """Recursively format the tree structure with proper tree lines."""
        lines = []
        items = list(structure.items())
        
        # Sort: directories first, then files, both alphabetically
        items.sort(key=lambda x: (x[1]['_type'] != 'tree', x[0].lower()))
        
        for i, (name, data) in enumerate(items):
            is_last = i == len(items) - 1
            
            # Determine the tree characters
            if is_root:
                current_prefix = ""
                next_prefix = ""
            else:
                current_prefix = prefix + ("└── " if is_last else "├── ")
                next_prefix = prefix + ("    " if is_last else "│   ")
            
            # Format current item
            if data['_type'] == 'tree':
                # Directory
                display_name = f"{name}/"
                if not is_root:
                    lines.append(f"{current_prefix}{display_name}")
                else:
                    lines.append(f"{display_name}")
                
                # Recursively add children
                children_output = self._format_tree_recursive(
                    data['_children'], 
                    next_prefix if not is_root else "", 
                    False
                )
                if children_output:
                    lines.append(children_output)
            else:
                # File
                size_info = f" ({data['_size']} bytes)" if data['_size'] > 0 else ""
                if not is_root:
                    lines.append(f"{current_prefix}{name}{size_info}")
                else:
                    lines.append(f"{name}{size_info}")
        
        return "\n".join(lines)
    
    def get_file_content(self, owner: str, repo: str, branch: str, file_path: str) -> str:
        """Get content of a specific file."""
        try:
            url = f"https://api.github.com/repos/{owner}/{repo}/contents/{file_path}?ref={branch}"
            response = requests.get(url, headers=self.headers, timeout=15)
            
            if response.status_code != 200:
                return f"❌ Error fetching file: HTTP {response.status_code}"
            
            file_data = response.json()
            
            if file_data.get('encoding') == 'base64':
                try:
                    content = base64.b64decode(file_data['content']).decode('utf-8')
                    return content
                except UnicodeDecodeError:
                    return "❌ Binary file - cannot display content"
            else:
                return file_data.get('content', 'No content available')
        
        except Exception as e:
            return f"❌ Error reading file: {str(e)}"
    
    def print_all_files(self, tree_data: Dict) -> str:
        """Print content of all files in the repository/subfolder."""
        if "error" in tree_data:
            return f"❌ {tree_data['error']}"
        
        tree = tree_data.get('tree', [])
        files = [item for item in tree if item['type'] == 'blob']
        
        if not files:
            return "📁 No files found to print."
        
        output = []
        output.append(f"📦 Repository: {tree_data['owner']}/{tree_data['repo']}")
        output.append(f"🌿 Branch: {tree_data['branch']}")
        if tree_data['subfolder']:
            output.append(f"📁 Subfolder: {tree_data['subfolder']}")
        output.append(f"📄 Total files: {len(files)}")
        output.append("\n" + "="*80 + "\n")
        
        sorted_files = sorted(files, key=lambda x: x['path'])
        
        for i, file_item in enumerate(sorted_files, 1):
            file_path = file_item['path']
            output.append(f"\n{'='*20} FILE {i}/{len(files)} {'='*20}")
            output.append(f"📄 Path: {file_path}")
            output.append(f"📊 Size: {file_item['size']} bytes")
            output.append("-" * 60)
            
            content = self.get_file_content(
                tree_data['owner'], 
                tree_data['repo'], 
                tree_data['branch'], 
                file_path
            )
            
            output.append(content)
            output.append("\n" + "-" * 60)
        
        return "\n".join(output)

# Initialize session state
if 'tree_data' not in st.session_state:
    st.session_state.tree_data = {}
if 'branches' not in st.session_state:
    st.session_state.branches = []
if 'tree_display' not in st.session_state:
    st.session_state.tree_display = ""
if 'files_content' not in st.session_state:
    st.session_state.files_content = ""

# Initialize analyzer
analyzer = GitHubRepoAnalyzer()

# Main title and description
st.title("🌳 GitHub Repository Tree Generator")
st.markdown("""
Generate directory trees and print file contents from any public GitHub repository.
Perfect for analyzing code structure and understanding project organization.
""")

# Sidebar for inputs
with st.sidebar:
    st.header("📝 Repository Configuration")
    
    # GitHub URL input
    github_url = st.text_input(
        "GitHub Repository URL",
        placeholder="https://github.com/owner/repository",
        help="Enter the full GitHub repository URL"
    )
    
    # Auto-fetch branches when URL changes
    if github_url and github_url.strip():
        try:
            owner, repo = analyzer.parse_github_url(github_url)
            with st.spinner("Fetching branches..."):
                branches = analyzer.get_branches(owner, repo)
                st.session_state.branches = branches
        except Exception as e:
            st.error(f"Invalid URL: {str(e)}")
            st.session_state.branches = []
    
    # Branch selection
    if st.session_state.branches:
        default_branch = 'main' if 'main' in st.session_state.branches else st.session_state.branches[0]
        try:
            default_index = st.session_state.branches.index(default_branch)
        except ValueError:
            default_index = 0
        
        selected_branch = st.selectbox(
            "Branch",
            options=st.session_state.branches,
            index=default_index,
            help="Select the branch to analyze"
        )
    else:
        selected_branch = st.text_input(
            "Branch",
            value="main",
            help="Enter branch name manually"
        )
    
    # Subfolder path
    subfolder_path = st.text_input(
        "Subfolder Path (optional)",
        placeholder="e.g., src/components",
        help="Leave empty to analyze the entire repository"
    )
    
    st.markdown("---")
    
    # Generate tree button
    if st.button("🌳 Generate Tree", type="primary", use_container_width=True):
        if not github_url.strip():
            st.error("Please enter a GitHub repository URL.")
        elif not selected_branch.strip():
            st.error("Please enter a branch name.")
        else:
            try:
                with st.spinner("Generating repository tree..."):
                    owner, repo = analyzer.parse_github_url(github_url)
                    subfolder = subfolder_path.strip()
                    
                    st.session_state.tree_data = analyzer.get_repo_tree(owner, repo, selected_branch, subfolder)
                    st.session_state.tree_display = analyzer.format_tree_display(st.session_state.tree_data)
                
                if "error" in st.session_state.tree_data:
                    st.error(st.session_state.tree_data["error"])
                else:
                    st.success("✅ Tree generated successfully!")
            
            except Exception as e:
                st.error(f"Error: {str(e)}")

# Main content area
col1, col2 = st.columns([1, 1])

with col1:
    st.header("📁 Repository Tree")
    
    if st.session_state.tree_display:
        # Display tree in a text area for easy copying
        st.text_area(
            "Tree Structure",
            value=st.session_state.tree_display,
            height=400,
            help="You can select and copy the text from this area"
        )
        
        # Action buttons
        col1a, col1b = st.columns(2)
        with col1a:
            st.download_button(
                label="💾 Download Tree",
                data=st.session_state.tree_display,
                file_name=f"{st.session_state.tree_data.get('owner', 'repo')}-{st.session_state.tree_data.get('repo', 'tree')}-tree.txt",
                mime="text/plain",
                use_container_width=True
            )
        with col1b:
            if st.button("📋 Copy Tree", use_container_width=True):
                st.code(st.session_state.tree_display, language="text")
                st.info("Tree structure displayed above - select and copy the text!")
    else:
        st.info("👈 Generate a tree using the sidebar to see the repository structure here.")

with col2:
    st.header("📄 File Contents")
    
    if st.session_state.tree_data:
        if st.button("📄 Print All Files", type="secondary", use_container_width=True):
            with st.spinner("Reading all files..."):
                st.session_state.files_content = analyzer.print_all_files(st.session_state.tree_data)
        
        if st.session_state.files_content:
            # Display files content in a text area for easy copying
            st.text_area(
                "Files Content",
                value=st.session_state.files_content,
                height=400,
                help="You can select and copy the text from this area"
            )
            
            # Action buttons
            col2a, col2b = st.columns(2)
            with col2a:
                st.download_button(
                    label="💾 Download Files",
                    data=st.session_state.files_content,
                    file_name=f"{st.session_state.tree_data.get('owner', 'repo')}-{st.session_state.tree_data.get('repo', 'files')}-content.txt",
                    mime="text/plain",
                    use_container_width=True
                )
            with col2b:
                if st.button("📋 Copy Files", use_container_width=True):
                    st.code(st.session_state.files_content, language="text")
                    st.info("Files content displayed above - select and copy the text!")
    else:
        st.info("Generate a tree first to print file contents.")

# Footer with instructions
st.markdown("---")
st.markdown("""
## 📋 Instructions:
1. **Enter GitHub URL**: Paste any public GitHub repository URL in the sidebar
2. **Select Branch**: Choose from auto-fetched branches (defaults to 'main')
3. **Subfolder (Optional)**: Specify a subfolder path to analyze only part of the repo
4. **Generate Tree**: Click to create the directory structure
5. **Print Files**: Click to display contents of all files in the tree
6. **Copy/Download**: Use the copy buttons or download buttons to save the results

## 🔧 Features:
- ✅ Auto-fetch available branches
- ✅ Support for subfolder analysis  
- ✅ File size information
- ✅ Organized tree structure
- ✅ Complete file content printing
- ✅ Copy and download functionality
- ✅ Error handling and validation
- ✅ Responsive layout

**Note**: For private repositories, you can add a `GITHUB_TOKEN` environment variable for authentication.
""")